Forecast Exchange Rate with Box Jenkins Model: An Empirical Study

Box and Jenkins (1970) introduced a family of linear stochastic models that are now referred as Box-Jenkins or ARIMA model. This paper practices the Box-Jenkins’ methodology, obtains the ARIMA models with British pound/U.S. dollar exchange rates. A comparison is given between forecast performance of...

Full description

Bibliographic Details
Main Author: Zheng, Jingwen
Format: Dissertation (University of Nottingham only)
Language:English
Published: 2012
Subjects:
Online Access:https://eprints.nottingham.ac.uk/25759/
Description
Summary:Box and Jenkins (1970) introduced a family of linear stochastic models that are now referred as Box-Jenkins or ARIMA model. This paper practices the Box-Jenkins’ methodology, obtains the ARIMA models with British pound/U.S. dollar exchange rates. A comparison is given between forecast performance of ARIMA models and the random walk model. The result suggests that the random walk model is superior to any of ARIMA models with the given data. Random walk is still the best technique method for exchange rate forecast since Meese and Rogoff (1983). GBP/USD exchange rate market efficiency holds.